Ultra-wideband (UWB) is a short-range wireless communication technology with strong resolution, precious detection and anti-jamming. UWB technology can complete the target recognition in the communication process. What's more, UWB has many great advantages in the aspect of transmission rate, power consumption and cost, so it is used in the personnel orientation, object detection and obstacle recognition widely. In this paper, the target recognition based on UWB is realized in rainy weather firstly with support vector machine (SVM). For the feature extraction, the method of principal component analysis (PCA) is used to reduce the dimension which can shorten the running time of the algorithm. In order to improve the recognition rate, the parameter of SVM should be optimized. Based on shuffled frog leaping algorithm (SFLA) and particle swarm optimization (PSO), this paper proposed PSO-SFLA which has better performance in global optimal ability and convergence speed.